Grant-less Uplink Transmission for LTE
Operated in Unlicensed Spectrum
Jinyu Zhang1, Wenting Chang2, Huaning Niu2 , Salvatore Talarico2, Hongwen Yang1
1
Beijing University of Posts and Telecommunications
2
Intel Mobile Communications Technology Ltd.
Email: zhangjinyu@bupt.edu.cn
arXiv:1802.05959v1 [eess.SP] 14 Feb 2018
Abstract—Deployment of Long Term Evolution (LTE)
in unlicensed spectrum has been a candidate feature to
meet the explosive growth of traffic demand since 3GPP
release 13. To further explore the advantage of unlicensed
bands, in this context the operation of both uplink and
downlink has been supported and studied in the subsequent
releases. However, it has been identified that scheduled
uplink transmission performance in unlicensed spectrum is
significantly degraded due to the double listen-before-talk
(LBT) requirements at both eNB when sending the uplink
grant, and at the scheduled UEs before transmission. In
this paper, in order to overcome this issue, a novel uplink
transmission scheme, which does not require any grant, is
proposed, and the details regarding the system design are
provided. By modeling the dynamics in time of the LBT for
both a system that employs a conventional uplink scheme,
as well as the proposed scheme, it is verified through an-
alytical evaluation that the double LBT scheme for uplink
transmission greatly reduces the channel access probability
for the UE, and leads consequently to performance loss,
while the proposed scheme is able to alleviate this issue.
System level simulation results, compliant with the LTE
standard, show that the proposed scheme can achieve a
significant performance gain in terms of throughput with
negligible performance loss for the downlink, and other
technologies operating in the same spectrum.
I. I NTRODUCTION
Owing the popularity of smartphones, tablets, and
other wireless devices, the recent widespread adoption of
wireless broadband has resulted in a tremendous growth
in the volume of mobile data traffic, which is projected
to continue unabated [1], [2]. As a consequence of this,
the system capacity of wireless communication systems
have been severely challenged. However, restricted by
the lack of available spectrum resource in licensed band,
the traditional Long Term Evolution (LTE) technology
is powerless in tackling this problem. Therefore, the
available resources in unlicensed band have attracted
recently more, and more attention as an important com-
plement to alleviate the high data traffic load [3], [4].
In this regards, 3GPP has introduced its operation in
unlicensed band via Licensed Assist Access (LAA) in
release 13 [5], [6]. LAA uses carrier aggregation in
the downlink to combine LTE in unlicensed spectrum
with LTE in the licensed band to expand the system
bandwidth. While significant changes have been made
compared to the LTE framework through the introduction
of several mechanisms [7], the authors of [8] have shown
that LAA ensures fair coexistence with existing Wi-Fi
networks.
Following the current momentum on unlicensed spec-
trum, recently 3GPP has started two new working items,
named “new radio (NR) based unlicensed access” [9]
and “Enhancements to LTE operation in unlicensed
spectrum” [10]. With this in mind, apart from LAA
systems, MulteFire (MF) systems [11] employ LTE
technology, but solely work in unlicensed spectrum
without assistance of the “anchor” in licensed spectrum.
For these systems the control information and reference
signal must be supported to be transmitted on unlicensed
carriers along with the entire data. In this regards, a
MF system is totally different than an LAA system, and
its system framework needs to be modified to support
the sole operation in unlicensed band. Even though
the MF technology is still at an embryonic stage, the
combination of LTE like performance benefits, and WiFi
like deployment simplicity makes MF a significantly
important supplement, and valuable study topic to meet
the ever-increasing wireless traffic.
In legacy LTE systems, scheduled based uplink (SUL)
transmission has been considered in the 3GPP release
14 study, wherein uplink transmission is conditional to
an explicit uplink grant via physical downlink control
channel (PDCCH) [12]. In order to comply with the FCC
regulation requirements, and in order to maintain fair
coexistence with other technologies, the listen-before-
talk (LBT) mechanism is applied to check whether the
channel is clear or occupied before using it. However,
the use of the legacy two stage modality for LBT reduces
the uplink channel access probability. This drawback
is highlighted and verified in terms of channel access
probability by modeling the dynamics in time of a
system employing LBT throughout a Markov chain,
similarly as [13]–[15]. Besides the penalty imputable to
a lowered channel access probability, the performance
of SUL is also negatively affect by the processing delay
(generally 4 ms due to hardware constraints) between the
uplink grant and the scheduled transmission, which may
also lead to transmission latency and resource waste in
case there is no downlink data. Hence, scheduled based
 First Partial SF
empty 1 2 3 4 5 6 7 8 9 10 11 12 13

MCOT
Fig. 1. Scheduled based uplink transmission mode

uplink transmissions are not suitable in unlicensed band.
In this paper, we consider a new uplink transmis-
sion scheme, which does not require any eNB schedul-
ing grant, named grant-less uplink (GUL) transmission.
While this methodology highly resembles that currently
used in the WiFi uplink design, it is a significant depar-
ture from the existing SUL transmission of the legacy
LTE. In this regards, a number of enhancements, which
are discussed along this manuscript, need to be made
with respect to the legacy LTE design in order to be
able to properly enable and perform GUL transmissions.
The rest of this paper is organized as follows. Sec-
tion II begins with a brief introduction of the SUL
scheme. This section continues by building an analytical
framework based on a Markov chain model, which is
employed to model the dynamics in time of the LBT
procedure for both the user equipment (UE) and eNB.
This analytical framework is then used to compare the
channel access probability of SUL and GUL schemes,
and highlight the benefits of the proposed scheme. The
overview procedure and design details for GUL mode
are then provided in section III. In section IV, the
performance of the proposed scheme is evaluated via
system level simulations. Finally, conclusions are drawn
in section V.
II. S YSTEM M ODEL
In legacy LTE, the UE that intends to transmit data
needs to obtain an uplink grant from the serving eNB,
and only then it can start uplink transmission, as il-
lustrated in Fig.1. In primis, the eNB is required to
perform Cat.4 LBT on the target carrier for the uplink
grant transmission, as regulated in [17], [18]. Once it is
able to access successfully the channel, the subsequent
maximum channel occupancy time (MCOT) can be
occupied, and scheduled for either downlink or uplink
transmission by the eNB. While the PDCCH carrying
uplink grant can be transmitted in the first available
subframe (SF), due to the processing delay the physical
uplink shared channel (PUSCH) is scheduled at the latter
SFs during the same MCOT. The remaining symbols
in the downlink SFs can be utilized for downlink data
transmission, if any. Before uplink transmissions can
take place, the scheduled UE needs to complete an
additional LBT (either single interval LBT or Cat.4 LBT)
[18] after receiving the grant. If this second LBT fails,
the resources reserved for uplink are wasted.
Intuitively, the SUL mode hampers the channel access
probability for the UE. In order to overcome this issue,
it is proposed here to adopt one-LBT uplink access
mechanism instead of this double-LBT procedure, which
leads to uplink transmissions that can be performed
autonomously without requiring grants, which we refer
to as GUL. For the proposed GUL scheme, on the other
hand, similarly to SUL, Cat.4 LBT is still employed for
the fair sharing of unlicensed band.
While Markov chains and their properties have been
extensively used to model and characterize the proce-
dure of LBT for Wi-Fi and LAA [13]–[16], in this
contribution they are used to model the LBT for MF
systems, in order to study its coexistence with Wi-Fi.
In particular, utilizing a Markov chain model, the LBT
procedure of a WiFi, and MF access node is modeled,
and the transmit probabilities in a randomly chosen slot
time can be calculated by (1) and (2), respectively [14],
[15]. For these equations, Q = 2(1−pb)(1−pf )(1−2pf ),

2q(1−pb)(1−2pf )
pW
tx
iF i
= (1)
2(1 − pb )(1 − pf )(1 − 2pf ) + q[W0 pf (1 − (2pf )m ) + (1 + W0 − 2pb )(1 − 2pf )]
2q(1−pb)(1−pf )R
pCat4
tx = (m+1)
(2)
Q+q[W0 P (1−pf )(1−(2pf ) )+P R(1−2pb)(1−2pf )+2R(1−pb)2 (1−pf )(1−2pf )]
 (m+1)
P = (pb + pf − pb pf ), R = (1 − pf ), q denotes
the probability of packet arrival, m is the maximum
clear channel assessment (CCA) stage, and W0 is the
initial contention window size. pf and pb denote the
probability of transmission failure due to collisions,
and the probability that the channel is detected to be
occupied, respectively. In [14], [15], these probabilities
are although determined under the simplified assumption
that all the nodes in the coexistence scenario can detect
the signal from all other nodes above the carrier sense
threshold. In order to address this issue, the distribution
of the detected energy [19]–[21] is here taken into
account. For simplicity, let assume that the path loss
between any two nodes are identical. The total receiving
power Prx can be then be obtained by multiplying the
receiving power P0rx from a single transmitting node
by the total number of the transmitters n. Thus, the
distribution of the energy detection conditioned on the
number of transmitters could be expressed as follows
(
1 y
µ y µ−1 e− 2 idle
fY (y|n) = 21 Γy(µ)µ−1 − 2γ+y √ (3)
2 ( 2γ )
2 e 2 Iµ−1 ( 2γy) busy
where γ is the signal-to-noise ratio (SNR), and de-
pends on the number of transmitters n since γ =
nP0rx /Pnoise , Γ (.) represents the gamma function, and
Iv (.) is the vth-order modified Bessel function of the
first kind.
Assume that the channel sense failure and the trans-
mission collision both occur in the case that the detected
energy is above the LBT threshold ythv . In a network
with N access nodes, it yields that
N −1  Z +∞
X N n
pb =pf = ptx (1−ptx)N−1−n fY (y|n) dy. (4)
n ythv
n=1
The transmission probability for a WiFi Access Point
(AP), and a system with Cat.4 LBT is evaluated then
by solving (1) and (2) using (4), respectively. For a SUL
scheme, only N eNBs perform Cat.4 LBT, and the uplink
channel is available only when both downlink Cat.4 LBT,
and the single slot LBT at the UE side succeed. Thus,
in this case the uplink channel access probability can be
expressed as
pSUL
tx = (1 − pb )pCat4
tx (5)
In the proposed GUL mode, the UE performs inde-
pendent Cat.4 LBT, which is nearly the same behaviour
as eNB in respect to the channel access procedure.
Therefore, the channel access probability for both UE
and eNB can be obtained by substituting within N in
(4) the sum of the number of UEs and eNBs involved.
Fig. 2 shows the channel access probability based on
the assumptions that ythv = −72 dBm for both the Wi-
Fi and the MF system. The number of WiFi APs or MF
eNBs deployed for each operator is N = 5, and each
eNB only has one active UE. For this plot, m = 4,
LW
&KDQQHODFFHVV SUREDELO\
: L) LLQ: L) L 6 8 / í 0 )
0 ) 8 / LQ: L) L 6 8 / í 0 )
: L) LLQ: L) L * 8 / í 0 )
0 ) LQ: L) L * 8 / í 0 )
3 UREDELO
LW
\ RI SDFNHWDUULYDO
Fig. 2. Channel access probability
W0 = 16, P0rx /Pnoise ≈ 10 and the following two
scenarios are shown:
• WiFi+SUL-MF: WiFi APs of operator 1 coexist
with MF eNBs and UEs of operator 2, which
operates in scheduled based uplink modality;
• WiFi+GUL-MF: WiFi APs of operator 1 coexist
with MF eNBs and UEs of operator 2, which
operates in grant-less based uplink modality.
As illustrated by Fig. 2, the SUL scheme is subject to a
small uplink channel access probability due to the double
LBT required, while for the GUL mode this improves
significantly with negligible impact on the WiFi system
performance.
III. S YSTEM D ESIGN OF G RANT- LESS U PLINK M ODE
Fig. 3 provides an illustration of the overall procedure
for the GUL mode. Firstly, the UE with uplink data
performs channel sensing. In this case, Cat.4 LBT is
adopted to maintain the fair coexistence with the incum-
bent system and other technologies. A preamble signal
is needed before data transmission for the detection at
the anchored eNB, and signalling of control information.
A reservation signal may also be needed to align with
the predefined boundary. Then, the UE can use the
whole MCOT for data transmission rather than shared
with downlink. Finally, the eNB needs to feedback the
ACK/NACK information for Hybrid Automatic Repeat
Request (HARQ) process.
The proper use of GUL mode needs a quite different
framework than that used by SUL in the LTE technol-
ogy. Therefore, a number of enhancements, such as the
control information and feedback, are needed with the
respect to the legacy LTE design, and the details are
discussed in this section.
A. Detection of PUSCH at eNB side
Due to the lack of scheduling, the serving eNB is not
aware of the UE’s transmission and it needs to detect
the presence of the uplink burst. Two candidate methods
can be taken into consideration for such indication:
 Partial
Synchronous
Super S
Asynchronou
Fig. 3. Grant-less uplink transmission mode
• Implicit indication by demodulation reference sig-
nal (DMRS): the serving eNB performs blind de-
tection of the DMRS sequence to infer the presence
of PUSCH;
• Explicit indication through Uplink Control Indicator
(UCI): in this context, the existing UCI formats
can be reused to provide additional information
regarding the uplink burst. The content of UCI
includes but is not limit to the following fields:
HARQ process number, UE identifier, and new data
indicator (NDI).
B. Uplink Sub-frame Design
Since the LBT could be completed at any time instant,
mostly not aligned with the primary cell (PCell) SF
boundary, this may result in a waste of resources due
to the fact that the transmission is postponed until the
boundary of next SF. In order to better utilize the interval
of time from the ends of the LBT until the PCell sub-
frame boundary, a more flexible design of the uplink SF
is required. As shown in Fig. 3, three uplink SF types
can be adopted:
• Synchronous SF, which is aligned with the bound-
ary of PCell SF to minimize the implementation
impact. In this context, a partial SF or super SF
can be defined on a subset of OFDM symbols
within the uplink SF (similar to the partial TTI
for downlink LAA), while the PCell still remains
aligned in terms of timing relationship with the
uplink burst transmission. In this case, the UE can
start PUSCH transmission at certain known OFDM
symbol positions within a SF with the aim to limit
the UE scheduling complexity. In particular, as
UE may know the duration of the partial TTI, it
may need to create multiple potential partial SFs
corresponding to different hypothesis of possible
partial SF. However, this incurs in a significant
computation and buffer complexity at the UE side.
Thus it is desirable to limit the set of possible
starting positions to assume some predefined and
restrict values, e.g. {1, 8}.
UL MCOT
• Asynchronous SF, which cannot be aligned with
PCell boundary as illustrated in Fig. 3. As long
as the channel is acquired through LBT, UE could
carry out the uplink transmission based on the
legacy 1 ms SF design.
C. Scheduling, Link Adaptation and HARQ Operation
Instead of relying on the indication from the serving
eNB, the UE needs to autonomously select the resource
allocation in GUL mode. Accurate channel state infor-
mation (CSI) is essential for both the scheduling at the
UE side, and the demodulation at the eNB side. Apart
from this, the UE also needs feedback information for
HARQ retransmission. In this regards, the process could
be summarized in following steps:
• Step 1: eNB estimates and calculates the uplink
CSI based on the sounding reference signals (SRSs)
from the UE. In particular, in this case the legacy
LTE design for SRS can be reused, and they can be
transmitted in the last OFDM symbol. Additionally,
the CSI request can be transmitted along with SRSs.
• Step 2: The UE chooses an appropriate modulation
and coding scheme (MCS). The selection can be
done by the eNB, which can indicate the best
suitable MCS to UE. Alternatively, the UE can
request CSI, and based upon this information it can
select the appropriate MCS by itself.
• Step 3: The UE transmits data along with the
scheduling information via PUCCH, which may
contain the HARQ process number and NDI.
• Step 4: The eNB transmits ACK/NACK feedback
via PDCCH, after receiving uplink data.
For link adaptation, the possible suitable options are:
• The eNB dynamically feedbacks the uplink CSI
for MCS selection, while indicating HARQ
ACK/NACK feedback;
• The UE uses the MCS indicated in latest DCI.
D. Control Channel Design
When UEs have simultaneous uplink data and control
transmission, control signaling can be multiplexed with
 Macro eNB
600 small cell eNB

400

200

−200

−400

−600

−800
−1000 −800 −600 −400 −200 0 200 400 600 800 1000

Fig. 5. Network layout for the indoor scenario

Fig. 4. Illustration of PUCCH control region

TABLE I
data prior to the discrete fourier transformation (DFT) M AIN S IMULATION PARAMETERS
to preserve the single-carrier property in the uplink
transmission as shown in Fig.4. This methodology can Parameters Value
Scenario Layout Indoor Scenario
be reused in systems such as MF, which work solely in
Number of UE 20 UEs per sector
unlicensed spectrum, but with certain extent. In fact, in Channel Model WINNER B+
these systems a different content can be carried, and a Carrier frequency 5GHz
list of possible fields is as follows: Inter-Station Distance 500m
MCOT 5ms
• Cell radio network temporary identifier (C-RNTI); Traffic Model FTP Model 3
• HARQ process number; File Size 0.5MByte
• NDI, which is used to state whether the current DL:UL Traffic Ratio 50:50
eNB Tx Power 18dBm
transmission is a retransmission or not; eNB Antenna Gain 5dB
• MCOT and uplink burst related information de- UE Tx Power 18dBm
scribed as the number of SFs. In case not all the SFs UE Antenna Gain 0dB
are used, the remaining SFs could be scheduled by
eNB for downlink or uplink transmission of other aim to evaluate the performance of the proposed scheme.
UEs. The simulations are performed under the assumptions
• Carrier used; agreed in [6], which are summarized in Table I. An
• A-CSI, and HARQ ACK/NACK bitmaps. indoor deployment with 7 hexagonal cell sites is con-
In order to reduce the UCI signalling overhead, sidered for each operator, as shown in Fig.5 . Every
it is preferred to transmit some of this information site has 3 sectors with 4 MF eNBs or 4 Wi-Fi APs
once, especially for fields such as A-CSI and HARQ randomly dropped, and grouped as a cluster. Similarly to
ACK/NACK bitmaps, while some other fields which are Fig. 2 two scenarios are considered: WiFi+SUL-MF, and
essential (i.e, HARQ process number, C-RNTI and NDI) WiFi+GUL-MF. Fig. 6 shows the performance in terms
should be transmitted in each SF. In this regard, at least of the mean user perceived throughput (UPT) for these
two different sizes for UCI should be predefined: one two scenarios for both a WiFi and MF systems and for
which includes the complete UCI with all fields, and both uplink and downlink.
one which incorporates only necessary fields. Since both This figure highlights that the uplink throughput of
the MCS index (which can be determined according to a MF operator is quite low when the SUL scheme
subsection III-C) and the UCI size are needed to separate is adopted for the aforementioned issues, and that the
the control information, the eNB can perform blind proposed GUL scheme allows to improve significantly its
detection to determine this last information. This option performance up to achieving comparable performances
is although very computational intense, and alternatively with WiFi. As for the downlink, the proposed scheme
the HARQ ACK/NACK or the rank indicator (RI) can leads to a slight performance loss, which is negligible
be used to indicate the UCI size. compared to the gain obtained for uplink. In fact, perfor-
mance in terms of sum throughput of both downlink and
IV. S IMULATION AND P ERFORMANCE E VALUATION uplink are still significantly improved with the proposed
This section provides the results obtained from com- scheme. Moreover, by comparing the performances of
prehensive system level simulations performed with the WiFi in terms of throughput between the case when
 45
40
35
30
UPT (Mbps)
25
20
15
10
5 LTE”.
0 [4] Qualcomm Whitepaper: “Extending the Benefits of LTE-A to
1 1.15 1.3 1.6
Total Offered Load for each UE (Mbps)
(a) Average uplink UPT
120
100
80
UPT (Mbps)
60
40
20
0 [9] 3GPP TD RP-170828, “New SID on NR-based Access to Unli-
1 1.15 1.3 1.6
Total Offered Load for each UE (Mbps)
(b) Average downlink UPT
Fig. 6. Average UPT performance
the SUL and the GUL scheme is used, it is possible to
notice that the proposed scheme is able to still guarantee
coexistence between MF and WiFi. For both downlink
and uplink, the WiFi performance is slightly degraded
due to the intense channel access competition with MF.
However, such performance degradation is acceptable,
and it may also incur in case the number of WiFi APs are
increased in a certain area. In conclusion, the proposed
GUL scheme can achieve remarkable performance gain
for MF uplink and maintain the friendly coexistence with
incumbent MF downlink, and Wi-Fi technology.
V. C ONCLUSION
In this paper, in order to cope with the tremendous
deterioration of the uplink performance of LTE systems
operating in unlicensed spectrum, such as MF, a new
transmission scheme is proposed, which allows to per-
form grant-less transmissions. By developing an analyt-
ical framework based on a Markov chain representation
of the LBT procedure, it shows that the GUL scheme is
able to increase the uplink channel access probability in
a MF system compared to a scheduled based scheme. In
addition, along the paper system designs and details on
how to enable this transmission scheme within the LTE
ecosystem are elaborated. Finally, comprehensive system
level simulations are provided, and evaluation indicates
that the proposed GUL mode can lead to a significant
improvement of the uplink UPT performance with the
negligible performance loss for MF downlink and Wi-Fi
systems.
REFERENCES
WiFi in WiFi+SUL-MF
WiFi in WiFi+GUL-MF [1] Cisco System, Cisco Visual Networking Index: GlobalMobile
MF in WiFi+SUL-MF Data Traffic Forecast Update, 2015-2020, [Online]. Available:
MF in WiFi+GUL-MF
https://www.cisco.com/c/en/us/solutions/collateral/service-
provider/visual-networking-index-vni/mobile-white-paper-c11-
520862.pdf
[2] Nokia Whitepaper: “Enhance mobile networks to deliver 1000
times more capacity by 2020”.
[3] Huawei Whitepaper: “U-LTE: Unlicensed Spectrum Utilization of
1.75 1.9 2.05 2.5 Unlicensed Spectrum”.
[5] 3GPP TD RP-141664,“Study on Licensed-Assisted Access using
LTE,” Sept. 2014.
[6] 3GPP TR 36.889 3rd Generation Partnership Project; Technical
WiFi in WiFi+SUL-MF Specification Group Radio Access Network; Study on Licensed-
WiFi in WiFi+GUL-MF
Assisted Access to Unlicensed Spectrum;(Release 13)[R]. 3GPP
MF in WiFi+SUL-MF
MF in WiFi+GUL-MF V13.0.0 (2015-06).
[7] A. Al-Dulaimi, S. Al-Rubaye, Q. Ni and E. Sousa,“5G Com-
munications Race: Pursuit of More Capacity Triggers LTE in
Unlicensed Band,” in IEEE Vehicular Technology Magazine, vol.
10, no. 1, pp. 43-51, March 2015.
[8] H. J. Kwon et al.,“Licensed-Assisted Access to Unlicensed Spec-
trum in LTE Release 13,” in IEEE Communications Magazine,
vol. 55, no. 2, pp. 201-207, February 2017.
1.75 1.9 2.05 2.5
censed Spectrum,” March 2017.
[10] 3GPP TD RP-170848, “New Work Item on Enhancements to LTE
operation in unlicensed spectrum,” March 2017.
[11] https://www.multefire.org/
[12] R. Karaki et al.,“Uplink Performance of Enhanced Licensed
Assisted Access (eLAA) in Unlicensed Spectrum,” 2017 IEEE
Wireless Communications and Networking Conference (WCNC),
San Francisco, CA, March 2017.
[13] G. Bianchi,“Performance analysis of the IEEE 802.11 distributed
coordination function,” in IEEE Journal on Selected Areas in
Communications, vol. 18, no. 3, pp. 535-547, March 2000.
[14] C. Chen, R. Ratasuk and A. Ghosh,“Downlink Performance
Analysis of LTE and WiFi Coexistence in Unlicensed Bands with a
Simple Listen-Before-Talk Scheme,” IEEE Vehicular Technology
Conference (VTC Spring), Glasgow, UK, May 2015.
[15] Y. Gao, X. Chu and J. Zhang,“Performance Analysis of LAA
and WiFi Coexistence in Unlicensed Spectrum Based on Markov
Chain,” IEEE Global Communications Conference (GLOBE-
COM), Washington, DC, Dec. 2016.
[16] Z. Fu, W. Xu, Z. Feng, X. Lin and J. Lin, ”Throughput Anal-
ysis of LTE-Licensed-Assisted Access Networks with Imperfect
Spectrum Sensing,” 2017 IEEE Wireless Communications and
Networking Conference (WCNC), San Francisco, CA, March
2017.
[17] ETSI EN 301 893 Broadband Radio Access Networks (BRAN);
5 GHz high performance RLAN; Harmonized EN covering the
essential requirements of article 3.2 of the R&TTE Directive[S].
ETSI V1.8.1 (2015-03).
[18] 3GPP TR 36.213 3rd Generation Partnership Project; Technical
Specification Group Radio Access Network; Evolved Universal
Terrestrial Radio Access (E-UTRA); Physical layer procedures
(Release 14). 3GPP V14.2.0 (2017-03).
[19] F. F. Digham, M. S. Alouini and M. K. Simon,“On the Energy
Detection of Unknown Signals Over Fading Channels,” in IEEE
Transactions on Communications, vol. 55, no. 1, pp. 21-24, Jan.
2007.
[20] F. F. Digham, M. S. Alouini and M. K. Simon, “On the energy
detection of unknown signals over fading channels,” , IEEE
International Conference on Communications (ICC), Anchorage,
AK, May 2003.
[21] B. Li, T. Zhang and Z. Zeng, “LBT with adaptive threshold for
coexistence of cellular and WLAN in unlicensed spectrum,” IEEE
International Conference on Wireless Communications & Signal
Processing (WCSP), Yangzhou, China, Oct. 2016.